// Query Grammar for Pest Parser
// This grammar supports the full query syntax
WHITESPACE = _{ " " | "\t" | "\n" | "\r" }
// Main entry point
query = { SOI ~ boolean_query ~ EOI }
// Boolean query with AND/OR operators (space-separated clauses default to implicit boolean)
boolean_query = { clause ~ (boolean_op ~ clause | clause)* }
clause = { required_clause | prohibited_clause | sub_clause }
required_clause = { "+" ~ sub_clause }
prohibited_clause = { "-" ~ sub_clause }
sub_clause = { grouped_query | field_query | term_query }
grouped_query = { "(" ~ boolean_query ~ ")" ~ boost? }
boolean_op = { ^"AND" | ^"OR" }
// Field-specific queries
field_query = { field ~ ":" ~ field_value }
field_value = { range_query | phrase_query | fuzzy_term | wildcard_term | simple_term }
// Range queries: [100 TO 500] or {A TO Z}
range_query = { range_inclusive | range_exclusive }
range_inclusive = { "[" ~ range_value ~ ^"TO" ~ range_value ~ "]" }
range_exclusive = { "{" ~ range_value ~ ^"TO" ~ range_value ~ "}" }
range_value = { "*" | quoted_string | unquoted_value }
unquoted_value = @{ (!WHITESPACE ~ !"TO" ~ !"]" ~ !"}" ~ ANY)+ }
// Phrase queries: "hello world" or "hello world"~10
phrase_query = { "\"" ~ phrase_content ~ "\"" ~ proximity? ~ boost? }
phrase_content = @{ (!"\"" ~ ANY)* }
proximity = { "~" ~ number }
// Fuzzy queries: term~2
fuzzy_term = { term ~ "~" ~ fuzziness? ~ boost? }
fuzziness = { number }
// Wildcard queries: te?t or test*
wildcard_term = { wildcard_pattern ~ boost? }
wildcard_pattern = @{
(ASCII_ALPHANUMERIC | "_")* ~ wildcard_char ~ (ASCII_ALPHANUMERIC | "_" | wildcard_char)*
}
wildcard_char = { "?" | "*" }
// Simple term with optional boost
simple_term = { term ~ boost? }
// Term query (standalone, no field)
term_query = { field_value }
// Boost: ^4 or ^0.5
boost = { "^" ~ boost_value }
boost_value = @{ number ~ ("." ~ number)? }
// Field name
field = @{ (ASCII_ALPHA | "_") ~ (ASCII_ALPHANUMERIC | "_" | ".")* }
// Term (word)
term = @{ escaped_char+ | unescaped_term }
unescaped_term = @{ (ASCII_ALPHANUMERIC | "_" | "-")+ }
escaped_char = @{ "\\" ~ special_char | normal_char }
special_char = { "+" | "-" | "!" | "(" | ")" | "{" | "}" | "[" | "]" | "^" | "\"" | "~" | "*" | "?" | ":" | "\\" | "/" }
normal_char = @{ ASCII_ALPHANUMERIC | "_" | "-" }
// Quoted string
quoted_string = { "\"" ~ quoted_content ~ "\"" }
quoted_content = @{ (!"\"" ~ ANY)* }
// Number
number = @{ ASCII_DIGIT+ }